INFECTION AND IMMUNITY, May 1992, p. 2008-20150019-9567/92/052008-08$02.00/0

Vol. 60, No. 5

Staphylococcus aureus and Derived Exotoxins Induce NuclearFactor KB-Like Activity in Murine Bone Marrow Macrophages

KLAUS BUSAM,1t CHRISTIANE GIERINGER,2:: MARINA FREUDENBERG,3AND HANS-PETER HOHMANN2t*

Institute ofBiochemistry, University of Freiburg, 1 and Max-Planck-Institute for Immunobiology,D-7800 Freiburg, Germany, and Central Research Units, Hoffmann-La Roche Ltd.,

Grenzacherstrasse 124, CH-4002 Basel, Switzerland2

Received 13 September 1991/Accepted 5 February 1992

Heat-killed gram-positive Staphylococcus aureus as well as S. aureus-derived exotoxins B and toxic shocksyndrome toxin 1 can induce nuclear factor cB (NF-KB)-like activity in murine bone marrow macrophages.The induction of NF-KB-like activity in murine macrophages by S. aureus was as effective as induction bytumor necrosis factor alpha (TNF-a) or lipopolysaccharides (LPS) and was observed in macrophages derivedfrom LPS-sensitive and LPS-resistant mice. Stimulation of macrophages with S. aureus but not with theexotoxins resulted in the accumulation of TNF-a in the culture medium. The induction of NF-cB-like activityby S. aureus, however, clearly preceded TNF-a secretion and was not inhibited by a neutralizing serum againstTNF-a. In addition, pretreatment of macrophages with the protein synthesis inhibitor cycloheximide ordexamethasone, which prevented the secretion of TNF-ce from macrophages, did not interfere with theinduction of NF-cB-like activity by S. aureus. This findings reveal the existence of bacterial components otherthan LPS which can induce NF-KB-like activity in susceptible cells.

Endogenous inflammatory mediators play an importantrole in the pathology caused by bacterial infections. Theprincipal toxins of gram-negative bacteria are lipopolysac-charides (LPS), which are constituents of the bacterial outermembrane (62). Many of the host responses to LPS can beattributed to macrophage-derived cytokines, such as tumornecrosis factor a (TNF-a) and interleukin-1 (IL-1) (49).Excessive amounts of these cytokines may lead to hypoten-sion, intravascular coagulation, multisystem organ failure,and even shock (47). Individual cytokines may mediatedifferent LPS effects. Thus, TNF-ot has been shown tomediate the lethality of LPS (6, 22, 23, 59) or tissue injury(4). IL-1 may activate lymphocytes (14). Anti-IL-1 andanti-TNF-a antisera inhibited certain effects of LPS or therespective cytokines in experimental animals (6, 15). Re-cently, it was recognized that, in addition to LPS, gram-negative bacteria contain at least one other common com-ponent which exhibits LPS-like biological activities (21).Although the interaction of macrophages with LPS is well

characterized, less is known about the components of gram-positive bacteria, e.g., Staphylococcus aureus, which wasused in the present study. These components are also potentactivators of macrophages. For example, a previous studyshowed that heat-killed S. aureus effectively stimulatesTNF-ot production in murine macrophages (21). Exotoxinssecreted from S. aureus were found to induce lymphocytefunction-associated molecule 1 in monocytes, B cells, andactivated T cells (48). Unlike LPS, S. aureus-derived exo-toxins, which have significant binding affinity for majorhistocompatibility complex (MHC) class II molecules (17,20, 46, 56), are strong T-cell mitogens. Complexes of S.

* Corresponding author.t Present address: Laboratory of Chemoprevention, National

Cancer Institute, Bethesda, MD 20892.t Present address: Vitamin Research Biotechnology, Hoff-

mann-La Roche Ltd., Grenzacherstrasse 124, CH-4002 Basel, Swit-zerland.

aureus-derived exotoxins and class II MHC molecules onantigen-presenting cells can interact with the T-cell recep-tors of a much larger proportion of T cells than can antigenicpeptides bound to class II MHC molecules (for a review, seereference 33).

In an attempt to characterize some of the mechanismswhich underlie the activation of macrophages by S. aureus,we studied the activation of nuclear factor KB (NF-KB) inthis event. NF-KB specifically binds to an 11-bp DNAfragment from the K light chain enhancer or to some slightvariants of that sequence (reviewed in references 27 and 40).NF-KB is involved in the transcriptional control of a varietyof genes, many of which are activated during the immuneresponse. NF-KB is involved in transcriptional activation ofthe K light chain, the IL-2 receptor a gene (7, 41), and genesencoding beta interferon (61) and TNF-a (11) and in theactivation of human immunodeficiency virus gene expres-sion (19, 28, 50). Potential NF-KB binding sites are alsopresent in the 5'-flanking regions of a number of other genes(61), in which they are also likely to be involved in thetranscriptional regulation of gene expression. NF-KB con-sists of two different subunits: a DNA binding protein of 51kDa and an associated 65-kDa protein (3, 36). Both subunitswere recently cloned and found to be homologous to theoncogene v-rel, its cellular homolog c-rel, and the Droso-phila maternal effect gene dorsal (8, 26, 37, 45, 51, 55). Thep65 subunit serves as a receptor for inhibitory protein IKB (3,51), which can prevent NF-KB from binding to the targetsequence (1). An IKB-like activity was cloned recently aswell (29). NF-KB is constitutively active in mature B cellsbut can be activated posttranslationally (57) in a variety ofother cell types. In unstimulated cells, NF-KB is located inan inactive form in the cytoplasm, in which it is complexedto IKB (2). Various inducers, such as phorbol esters, LPS,double-stranded RNA, and certain viruses, allow NF-KB tobe released from the complex and to travel to the nucleus.NF-KB can also be activated by the cytokines TNF-a,TNF-1, and IL-1 (30, 50), suggesting a mechanism for

2008

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

NF-KB ACTIVATION BY S. AUREUS 2009

autostimulatory and costimulatory regulatory loops (40).Prolonged activation of NF-KB observed in LPS- and TNF-a-stimulated cells requires de novo protein synthesis (31)and is committed by enhanced levels of mRNA encoding thepS1 DNA binding subunit of NF-KB (45).

Since components of S. aureus are some of the mostpotent activators of macrophages and NF-KB seems to be avital control element in the process leading to the activationof these cells, we addressed the role of NF-KB in macro-phages exposed to whole S. aureus and S. aureus-derivedexotoxins.

MATERIALS AND METHODS

Mice. The LPS-responder C57BL/lOScSn and the LPS-nonresponder C57BL/lOScCr mice (12, 21) of both sexes

were obtained from the breeding stock of the Max-Planck-Institut. Six-week-old animals served as donors of bonemarrow cells.

Cells. Bone marrow cells were flushed from mouse femoraand cultivated at a concentration of 5 x 105 cells per ml inhydrophobic Teflon film bags (Heraeus, Hanau, Germany)as described previously (18). The culture medium consistedof 70% high-glucose-level Dulbecco modified Eagle medium(GIBCO BRL GmbH, Karlsruhe, Germany) containing 10%fetal calf serum, 5% horse serum, 0.01 mM sodium pyruvate(GIBCO), 50 nM 2-mercaptoethanol (Roth, Karlsruhe, Ger-many), 50 U of penicillin, and 50 ,ug of streptomycin (Se-romed, Berlin, Germany) per ml and 30% L-cell-conditionedmedium (44) and was prepared as described previously (22).After 10 days of culturing, the cells were harvested bycentrifugation, washed twice with phosphate-buffered saline(PBS), and resuspended in Dulbecco modified Eagle mediumsupplemented with 1% (vol/vol) fetal bovine serum at a

concentration of 106 cells per ml. Ten milliliters of cellsuspension was used per dish (8.6-cm diameter). After 24 h,the adherent cells were washed once and incubated withmedia containing various stimuli. Human promyelocyticHL60 cells were cultivated and stimulated with TNF-a as

described previously (30).Bacteria. Preparations of heat-killed S. aureus were ob-

tained as described previously (21). Macrophages were

stimulated with 20 to 50 ,ug of heat-killed bacteria per ml ofculture medium.

Materials. LPS from Salmonella minnesota R595 (roughform) was isolated, purified, and converted to the uniformtriethylamine salt form as described previously (24, 25).Recombinant murine TNF-a (1.2 x 1010 U/mg) was a giftfrom G. R. Adolf (Boehringer Institut fur Arzneimittelfors-chung, Vienna, Austria). S. aureus-derived exotoxins B andtoxic shock syndrome toxin 1 (TSST-1) were obtained fromToxin Technology Inc., Madison, Wis. Unless indicatedotherwise, the macrophages were stimulated with 500 ng ofLPS, 100 ng of TNF-a, and 70 pg of exotoxins per ml ofculture medium. Rabbit serum raised against recombinantmurine TNF-ot was purchased from Genzyme Corp., Bos-ton, Mass., and used at a 1:1,000 dilution. Dexamethasone(Sigma) and cycloheximide (Fluka, Buchs, Switzerland)were used at 1 ,uM and 100 ptg/ml, respectively.

Preparation of nuclear extracts and EMSA. The mediacontaining the stimulating agents were removed from themacrophages, the cells were detached from the culturedishes after 5 min of incubation at 37°C with PBS containing1 mM EDTA and collected by centrifugation, and about 2 x

107 cells were resuspended in 500 ,ul of hypotonic lysis bufferA (13). After 20 min, the cells were homogenized by 20

strokes with a loose-fitting Dounce homogenizer. The cellhomogenates were centrifuged for 4 min at 6,500 rpm in aMicrofuge (ca. 4,000 x g). The supernatants (6.5k superna-tants) were removed, and the pellets were extracted withfour packed pellet volumes of high-salt buffer (buffer C) (13).After 60 min, the samples were centrifuged as describedabove. The high-salt extracts were diluted with 3 volumes oflow-salt buffer (buffer D) (13) containing 1% Nonidet P-40and were used immediately for electrophoretic mobility shiftassays (EMSA) or kept frozen at -20°C. EMSA wereperformed as described previously (58). Five thousandcounts per minute of a 32P-end-labeled DdeI-HaeIII frag-ment of the K light chain enhancer containing the NF-KBbinding site was used per assay. A restriction fragmentmutated in the NF-KB binding site (aattaAC'TTCC insteadof GGGGAC(TTTCC) but otherwise identical to the wild-type fragment was used as a negative control. This mutatedfragment does not bind NF-i-B (39). In some experiments, aprotease inhibitor cocktail consisting of 10 mM benzamidine,100 U of aprotinin per ml, 10,uM leupeptin, 1,M pepstatin,1 mM o-phenanthroline, and 1 mM phenylmethylsulfonylfluoride was present during nuclear extract preparation. Theprotease inhibitors were purchased from Fluka.

Methylation interference. Methylation interference assayswere performed as described previously (31).Measurement of TNF-a activity. TNF-a accumulation in

macrophage supematants was determined with an L929 cellcytotoxicity assay as described previously (21). The detec-tion limit was 10 pg/ml.

RESULTS

S. aureus induces NF-KB-like proteins in murine macro-phages. Bone marrow macrophages from LPS-resistantC57BL/lOScSr mice (CR macrophages) and LPS-sensitiveC57BL/lOScSn mice (SN macrophages) were incubated withS. aureus (Fig. 1, lanes 3, 5, 9, and 12). Activated KB bindingfactors in nuclear extracts of these cells were detected byEMSA with an oligonucleotide containing the intact NF-KBbinding site of the K light chain enhancer. To control fornonspecifically binding proteins, we used a fragment mu-tated in the NF-KB binding site but otherwise identical to thewild-type fragment. In addition, HL60 cells (Fig. 1, lanes 15and 17) and CR and SN macrophages (Fig. 1, lanes 4, 6, 10,and 13) were stimulated with TNF-a, a well-known activatorof transcription factor NF-KB (52). Parallel cell cultures wereleft untreated (Fig. 1, lanes 1, 7, 14, and 16). With nuclearextracts of TNF-ct- and S. aureus-stimulated CR and SNmacrophages, two retarded complexes, complex 1 and com-plex 2, were observed. Complex 1 comigrated with theauthentic NF-KB-DNA complex (45) from TNF-a-stimu-lated HL60 cells, and complex 2 migrated somewhat faster.Both complexes were virtually undetectable in unstimulatedcells and were not generated with the mutant DNA fragment.A band located between complex 2 and the free oligonucle-otide was considered to be unrelated to NF-KB, because itwas detected with the wild-type oligonucleotide and waseven more pronounced with the mutant oligonucleotide.Neither complex 1 nor complex 2 was observed with nuclearextracts of LPS-treated CR macrophages (Fig. 1, lane 2).They were present, however, in nuclear extracts of LPS-treated SN macrophages (Fig. 1, lane 8), confirming theresistance of CR macrophages to LPS. These results indicatethat DNA binding proteins which were similar or identical toNF-KB (see below) were induced upon S. aureus adminis-tration in mouse bone marrow macrophages. Since LPS-

VOL. 60, 1992

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2010 BUSAM ET AL.

-Wildtype- Mutant -Wildtype- Mutant Wildtype Mutant1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 -17

-- * .A1

L.,-w....

C2--

0iP1 0 ri. <

NF-KB

F-W UJ vJ 0 VJ f_

0 _.1 T Z (1 Z

4-Q 4u..cn Uc

s: X :5_. .

c3co3

Qf LL_. a~z

cc

- LPS-resistant CR-MGD- - LPS-sensitive SN-MDA-

0 Z 0 Zc c ~

HL60 cells

FIG. 1. S. aureus activation of NF-KB-like DNA binding proteins in murine macrophages. Macrophages from LPS-resistant (CR-Mb) andLPS-sensitive (SN-M4) mice were stimulated for 1 h at 37°C with LPS and S. aureus. In addition, the macrophages and HL60 cells werestimulated with TNF-a under similar conditions. Parallel cell cultures were left untreated. NF-KB(-like) activity was measured by EMSA withan oligonucleotide that contained the wild-type binding site for NF-KB from the K light chain (58) (Wildtype) or, as a control, an inactivemutated binding site that does not bind NF-KB (39) (Mutant). Nuclear extracts obtained from aliquots of each cell type were used. Thepositions of specifically retarded complexes 1 and 2 (Cl and C2) from activated macrophages, of NF-KB from HL60 cells, and of theunretarded DNA fragment (F) are indicated.

induced activation of the NF-KB-like proteins was observedin SN but not CR macrophages, the possibility that LPScontamination in the reagents or cell preparations used wasresponsible for NF-KB activation upon stimulation with S.aureus could be excluded.

S. aureus-induced NF-KB-like proteins from mouse macro-phages are very similar to NF-cB from TNF-cx-stimulatedhuman HL60 cells. In the experiment shown in Fig. 1,nuclear extracts were prepared in the presence of severalprotease inhibitors to reduce the activity of endogenousproteases. When the inhibitors were omitted, neither com-plex 1 nor complex 2 was generated with nuclear extracts ofS. aureus-stimulated (Fig. 2A, lane 1) or TNF-ot-stimulated(Fig. 2B, lane 1) CR macrophages. Instead, a faster-migrat-ing complex, complex 3, was observed. When nuclearextracts were prepared from a mixture of TNF-ao-stimulatedCR macrophages and TNF-a-stimulated HL60 cells in theabsence of protease inhibitors, only complex 3 was obtained(Fig. 2B, lane 2), whereas intact NF-KB activity was ob-served with nuclear extracts prepared from TNF-a-stimu-lated HL60 cells in the absence of macrophages (Fig. 2B,lane 3). The formation of complex 3 was not detected withthe mutant oligonucleotide (Fig. 2B, lanes 4 to 6). Theseresults reveal the presence in macrophages but not HL60cells of proteases which can degrade NF-KB. A similar stableproteolytic fragment of NF-KB visualized as complex 3 fromS. aureus-stimulated macrophages and TNF-a-stimulated

HL60 cells was generated and specifically bound to the DNAfragment. To test whether complex 2, which was observed inaddition to complex 1 with protease inhibitor-treated nuclearextracts from mouse macrophages, might result from incom-plete protease inhibition, we incubated extracts of TNF-ao-stimulated HL60 cells (Fig. 3C, lane 5) with nuclear extractsof unstimulated mouse macrophages (Fig. 3C, lanes 1 and 3)and reacted them with the wild-type oligonucleotide. In theabsence of the protease inhibitor cocktail, only complex 3was observed (Fig. 3C, lane 2), whereas in the presence ofthe protease inhibitor cocktail, only the complex of intactNF-KB and the wild-type DNA fragment could be detected(Fig. 3C, lane 4). Thus, the protease inhibitor cocktailcompletely inhibited the in vitro degradation of NF-KB bymouse macrophage proteases. Methylation interference as-says with the upper strand of the DNA fragment containingthe NF-KB binding motif revealed identical patterns ofguanine contact points for NF-KB from HL60 cells and themacrophage proteolytic fragment extracted from complex 3(Fig. 2D). In summary, protease clipping and methylationinterference confirmed the similarity between NF-KB fromHL60 cells and the macrophage KB binding factor. Hereaf-ter, we call both factors NF-KB for the sake of simplicity,although we are aware that the experiments described abovedo not finally prove their identity.

S. aureus-induced NF-KB activation is not mediated byTNF-ao TNF-ax is released from mouse macrophages upon

INFECT. IMMUN.

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

NF-KB ACTIVATION BY S. AUREUS 2011

( ) 12 3 4 (i) 1 2 3 4 5 6

NF-icB-

FF-

:3-o =3

= E= Eno yes

prot. inhibitors

O 1 2345._DS000

NF-K;Bi

C3- 0

F

2 2-@ iu>+X1 +=:o o-i -JI

-wildt.- - mut.-

w iTC

C

..-::r

R F RF

dP qpeKo4 .I 1>eo

+ :to

oCD CD

-J -J

I I

-no--yes-noprot. inhibitors

FIG. 2. Similar proteolytic fragments and identical methylation interference patterns obtained from S. aureus-activated NF-KB-like DNAbinding proteins and authentic NF-KB. (A) Nuclear extracts of murine macrophages stimulated for 1 h at 37°C with S. aureus were preparedin the absence or presence of various protease inhibitors (prot. inhibitors). (B) Murine CR macrophages (MbI), HL60 cells, and mixtures ofthese cell types were stimulated with TNF-a, and nuclear extracts were prepared in the absence of protease inhibitors. (C) Nuclear extracts(5 p.1) from TNF-a-stimulated HL60 cells were mixed with 3 ,ul of the 6.5k supernatants of mouse macrophage homogenates prepared in theabsence (lane 2) or presence (lane 4) of the protease inhibitor cocktail or with 3 ,ul of buffer D (lane 5). In addition, 3 ,ul of the 6.5k supernatantsof mouse macrophage homogenates were mixed with 5 Il of buffer D (lanes 1 and 3). The mixtures were incubated for 1 h at 37C.NF-KB(-like) activities were measured in the extracts by EMSA. Wild-type (wildt.) and mutant (mut.) DNA fragments were used. Thepositions of complexes 1, 2, and 3 (Cl, C2, and C3), of NF-KB, and of the unretarded DNA fragment (F) are indicated. (D) Methylationinterference assays were performed as described previously (36). Nuclear extracts of S. aureus-stimulated murine CR macrophages (Mt) orpartially purified NF-KB from TNF-a-treated HL60 cells prepared as described by Kawakami et al. (36) were subjected to EMSA with apartially methylated DNA fragment. Pieces of the EMSA gels containing the free DNA fragment, the DNA-NF-KB complex, or the complexofDNA and the proteolytic fragment were cut out, and the DNA was extracted and chemically degraded. Reaction products of the free DNA(F) and the DNA isolated from the retarded DNA-protein complexes (R) were applied to a sequencing gel. The positions of G residues which,upon methylation, interfered with protein binding are indicated by arrowheads.

stimulation with S. aureus (21). Since TNF-a can activateNF-KB in macrophages (Fig. 1), S. aureus-induced NF-KBactivation might occur by an autocrine mechanism followingTNF-a secretion. To test this possibility, we stimulatedmouse macrophages with S. aureus in the presence orabsence of an anti-TNF-a serum. At the dilution used, theantiserum was able to prevent NF-sB activation in macro-phages after the administration of 1 ng of murine recombi-nant TNF-a per ml (Fig. 3A, lanes 2 and 3). This TNF-aconcentration was much higher than the concentration thatcould have maximaly accumulated within 1 h of stimulationof the macrophages with S. aureus (see below). The anti-TNF-a serum did not interfere with S. aureus-inducedNF-KB activation (Fig. 3A, lanes 4 and 5). Furthermore, thedependence of S. aureus-induced NF-KB activation on denovo protein synthesis was studied with cycloheximide (Fig.3B). Cycloheximide induced NF-KB activity in mouse mac-rophages (Fig. 3B, lane 2). In nuclear extracts of macro-phages treated with cycloheximide and then with S. aureus,enhanced levels of active NF-KB were observed, comparedwith those in nuclear extracts of macrophages treated withcycloheximide or S. aureus only (Fig. 3B, lanes 2 to 4). This

result suggests that NF-KB activation by S. aureus did notrequire de novo protein synthesis. In addition, S. aureus wasapplied in the presence of dexamethasone, a well-knowninhibitor of cytokine production (5, 10, 23, 35). Dexametha-sone added to bone marrow macrophages 1 h prior to theadministration of S. aureus had no effect on NF-KB activa-tion (Fig. 3C). The amount of TNF-a, however, that wassecreted into the culture supematant of the macrophageswithin 6 h after the administration of S. aureus was substan-tially reduced by dexamethasone (from 3.8 to 0.1 ng/ml).Finally, whereas maximal NF-KB activation in murine bonemarrow macrophages was achieved by S. aureus treatmentin less than 1 h (Fig. 3D), no TNF-a was detectable in thecell supernatant at this time.

Bacterial exotoxins contribute to S. aureus-induced NF-KBactivation. To extend our understanding of the interactionbetween S. aureus and NF-KB, we also studied macrophageactivation by bacterial exotoxins. S. aureus-derived exotox-ins are very potent activators of human and murine leuco-cytes, inducing the release of large amounts of cytokines,such as IL-2, TNF-a, and TNF-p (see Discussion). SinceNF-KB is thought to be involved in the regulation of the

VOL. 60, 1992

. ... K

1 2 3 4-M

jj4,.,-;.- Itm

In,

Ci-

C2-C3

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2012 BUSAM ET AL.

1 2 3 4 5A._b .^io,

0_® -Wildtype-

1 2 3 4-Mutant -5 6 7 8 1 2 3 4

O - Wildtype- Mutant1 2 3 4 5 6 7 8

C2-

0::3-

i3ai

F

CLO C>o 00 C\J L) (6o 6

Incubation time withSt. aureus (hours)

#) Y (i)

a LL L DDO z z S0Z§Z n

<; (n U

(I)cl)

CD x< 'D X< WC D a( D0 -= (]c "2

V :n4-; Q C.

C/) + c/)

(iCD<1L?VWCDx

3X s

h

ozOBC a 0 S 0

$OC= eCZ )c

FIG. 3. S. aureus-induced NF-KB activation is not mediated by TNF-a. (A) Murine LPS-resistant macrophages were stimulated for 1 hwith 1 ng of murine TNF-a per ml (lanes 2 and 3) or with 20 ,ug of S. aureus (lanes 4 and 5) in the absence or presence of 1:1,000-dilutedanti-TNF-a serum (AS). (B) Murine LPS-resistant macrophages were treated for 1 h with cycloheximide (cyclohex.), S. aureus was added,and stimulation was continued for an additional hour. (C) Murine LPS-resistant macrophages were treated for 1 h with dexamethasone andthen stiumlated with an S. aureus suspension for 6 h. From these cells and from the cells mentioned in panels A and B, nuclear extracts wereprepared in the presence of protease inhibitors. (D) Murine LPS-resistant macrophages were stimulated for the indicated periods of time withS. aureus, and nuclear extracts were prepared in the absence of protease inhibitors. The nuclear extracts were analyzed by EMSA with an

oligonucleotide containing the wild-type binding site for NF-KB (Wildtype). In panels B and D, the inactive mutated binding site (Mutant) wasused as well. The positions of complexes 1, 2, and 3 (Cl, C2, and C3) and of the unretarded DNA fragment (F) are indicated.

expression of the genes encoding these factors, we investi-gated whether the S. aureus-derived exotoxins could lead toNF-Kd activation in murine macrophages. Within 1 h ofstimulation with exotoxin B and TSST-1, NF-KB activationcomparable to that obtained with whole heat-killed bacteriawas observed (Fig. 4, lanes 2, 3, and 4). TNF-a, activity inthe culture supernatant of murine macrophages treated for 6or 24 h with exotoxin B or TSST-1 was below the detectionlimit (>10 pg/ml), whereas massive secretion of TNF-aactivity was observed upon stimulation with whole heat-killed bacteria (3.7 ng/ml after 6 h). Thus, the tested bacterialexotoxins represent one class of S. aureus components thatcan induce NF-KB activation. Stimulation of macrophageswith these exotoxins, however, is not sufficient to induceTNF-a secretion.

DISCUSSION

In this study, we demonstrated that S. aureus activatedDNA binding proteins in murine bone marrow macrophagesthat bound to a DNA fragment containing the NF-KB motifof the K light chain enhancer, giving rise to three differentDNA-protein complexes. According to the binding specific-ities, native proteolytic fingerprints, and methylation inter-ference patterns, these DNA binding proteins should be

closely related to NF-KB. The most retarded complex,complex 1, had an electrophoretic mobility in EMSA gelssimilar to that of the complex between the DNA fragmentand authentic NF-KB. The protein giving rise to complex 2,which moved faster than complex 1 in EMSA gels, shouldalso be related to NF-KB, as indicated by the bindingspecificity and activation profile of the protein. Complex 1and complex 2 were converted to complex 3, whichwas leastretarded in EMSA gels, by one or several macrophage-specific proteases, suggesting that the protein components incomplexes 1 and 2 share similar structures. Whereas theprotein component of complex 1 should be identical to intactmouse NF-KIB, complex 2 might represent an NF-KB speciesthat has an altered subunit composition. Alternatively, theprotein component of complex 2 might be a partially de-graded form of NF-KB. Such a putative degradation ofNF-KB, however, would have to have occurred in intactcells, since no proteolytic activity degrading NF-KB was

detected in macrophage extracts after they had been treatedwith the protease inhibitor cocktail. A time course deter-mined for the in vitro degradation of NF-KB by macrophageproteases in the absence of any protease inhibitor did notindicate that the protein component of complex 2 was an

intermediate of the stable degraded NF-KB species presentin complex 3 (data not shown). A similar heterogeneity in

Cl-

C2-

C1-

C2-

F- F

0

I

F-I

INFECT. IMMUN.

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

NF-KB ACTIVATION BY S. AUREUS 2013

-Wildtype - -Mutant-1 2 3 4 5 6 7 8

Cl-

C2-

F-

U

J( F C

FIG. 4. Bacterial exotoxins B and TSST-1 are activators ofNF-KB. Murine LPS-resistant macrophages were stimulated for 1 hwith the S. aureus-derived exotoxins B and TSST-1. Parallel cellcultures were treated with S. aureus or left untreated (none). NF-KBactivity was measured in nuclear extracts of the cells by EMSA withan oligonucleotide containing the wild-type binding site for NF-KB(Wildtype) or the inactive mutated binding site (Mutant). Thepositions of specifically retarded complexes 1 and 2 (Cl and C2) andof the unretarded DNA fragment (F) are indicated.

NF-KB binding activities was found in nuclear extracts frommouse macrophages in two recent studies (11, 28). In sum-

mary, DNA binding proteins which were very similar oreven identical to NF-KB were activated in murine macro-

phages upon stimulation with S. aureus.

Since NF-KB takes part in the signal pathway of TNF-aand TNF-ax is produced by macrophages after exposure to S.aureus, we had to rule out the possibility that TNF-amediated S. aureus-induced activation of NF-KB. Four linesof evidence argue against TNF-a as a mediator. First,NF-KB activation was induced in the presence of dexam-ethasone, which inhibited TNF-a production. Second, ananti-TNF-o serum did not interfere with S. aureus-inducedNF-KB activation. Third, S. aureus-induced NF-KB activa-tion was enhanced in the presence of cycloheximide atconcentrations inhibiting de novo protein synthesis. Cyclo-heximide-induced NF-KB activation was also reported byother investigators (11, 57). Fourth, NF-KB activation oc-curred much faster than TNF-a secretion. It should also benoted that NF-KB activation but not TNF-a secretion was

observed after stimulation of murine macrophages withexotoxin B and TSST-1.CR macrophages cannot activate NF-KB in response to

LPS, whereas NF-KB activation readily occurs upon stimu-lation with S. aureus and TNF-ac. This finding confirms theLPS resistance of CR macrophages, but it also demonstratesthat LPS resistance is manifested early in the signal pathway

of LPS, before common components are shared with theNF-KB induction pathways used by S. aureus and TNF-(x.Mutant strains of the mouse pre-B-cell line 70Z/3 failed toexpress the K light chain and were unable to activate NF-KBupon stimulation with LPS, IL-1, and 12-O-tetrade-canoylphorbol-13-acetate (9). Therefore, signal transductionin these variant strains should be affected at a step commonto these three stimulators.

S. aureus exotoxins bound to class II MHC molecules (17,20, 46, 56) on accessory cells, e.g., monocytes, induced Tcells to proliferate and to secrete cytokines, such as IL-2,gamma interferon, and TNF-a (16, 54, 60). It is a matter fordebate, however, whether monocytes or macrophages canbe triggered directly by the exotoxins to secrete TNF-a orTNF-P or whether the cytokine secretion that was observedin cultures of monocytes isolated from peripheral bloodleucocytes (32, 34, 53) might have depended on a minorfraction of contaminating T cells possibly present in thesecultures. Our data, which are in accordance with the resultsof Fischer and collegues (16), indicated that exotoxin B andTSST-1 did not induce TNF-a secretion from in vitro-differentiated murine macrophages. Nevertheless, the inter-action of these exotoxins with these cells gave rise to anintracellular signal leading to NF-KB activation. It might beconcluded from these results that class II MHC moleculespresent on the surface of macrophages have, in addition totheir antigen-presenting function, a direct function in intra-cellular signaling. Accordingly, Mourad and colleaguesshowed that the binding of TSST-1 to class II MHC mole-cules activated lymphocyte function-associated molecule 1in monocytes, B cells, and T cells and suggested a new rolefor class II MHC molecules as signal-transducing receptors(48). However, we cannot exclude the possibility that thereare additional cell surface receptors for exotoxin B andTSST-1 on murine macrophages that can act as signal-transducing receptors mediating NF-KB activation indepen-dently from class II MHC molecules.The components of S. aureus that are responsible for the

effects of whole bacteria on murine macrophages should bedifferent from exotoxin B and TSST-1, since (i) S. aureusexotoxins are secreted into the surrounding medium andtherefore should not accumulate on the bacterial surface,and (ii) S. aureus was subjected to rigorous heat treatmentprior to use. Proteins such as exotoxins should becomedenatured under these conditions. We are currently investi-gating the precise chemical nature of the S. aureus compo-nents that stimulate NF-KB activation and TNF-a secretionin murine bone marrow macrophages. In addition, we aretrying to elucidate the role of class II MHC molecules in thesignal transduction of the bacterial exotoxins.

ACKNOWLEDGMENTSWe thank H. Stubig and C. Steidle for excellent technical assis-

tance, C. Scheidereit for performing the methylation interferenceexperiment, C. Galanos and A. P. G. M. van Loon for criticalreading of the manuscript, and K. Decker for continuous support.

This work was supported in part by grants from the Fond derChemischen Industrie, from the Deutsche Forschungsgemeinschaftthrough SFB 154, and from the Bundesminister fur Forschung undTechnologie (grant 01 Ki 8809).

REFERENCES1. Baeuerle, P. A., and D. Baltimore. 1988. Activation of DNA-

binding activity in an apparently cytoplasmic precursor of theNF-KB transcription factor. Cell 53:211-217.

2. Baeuerle, P. A., and D. Baltimore. 1988. IKB: a specific inhibitorof the NF-KB transcription factor. Science 242:540-554.

VOL. 60, 1992

I- yp-%

a? :: k."

1. p .AF

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2014 BUSAM ET AL.

3. Baeuerle, P. A., and D. Baltimore. 1989. A 65-kD subunit ofactive NF-KB is required for inhibition of NF-KB by IKB. GenesDev. 3:1689-1698.

4. Beutler, B., and A. Cerami. 1986. Cachectin and tumour necro-sis factor as two sides of the same biological coin. Nature(London) 320:584-588.

5. Beutler, B., and A. Cerami. 1989. The biology of cachectin/TNF-a primary mediator of the host response. Annu. Rev.Immunol. 7:625-655.

6. Beutler, B., I. W. Milsark, and A. C. Cerami. 1985. Passiveimmunization against cachectin/tumor necrosis factor protectsmice from lethal effect of endotoxin. Nature (London) 320:869-871.

7. Bohnlein, E., J. W. Lowenthal, M. Siekevitz, D. W. Ballard,B. R. Franza, and W. C. Greene. 1988. The same induciblenuclear proteins regulate mitogen activation of both the inter-leukin-2 receptor-alpha gene and type 1 HIV. Cell 53:827-836.

8. Bours, V., J. Villalobos, P. R. Burd, K. Kelly, and U. Siebenlist.1990. Cloning of a mitogen-inducible gene encoding a KB DNAbinding protein with homology to the rel oncogene and tocell-cycle motifs. Nature (London) 348:76-80.

9. Briskin, M., M. Damore, R. Law, G. Lee, P. W. Kincade, C. H.Sibley, M. Kuehl, and R. Wall. 1990. Lipopolysaccharide-unresponsive mutant pre-B-cell lines blocked in NF-KB activa-tion. Mol. Cell. Biol. 10:422-425.

10. Busam, K. J., T. M. Bauer, J. Bauer, W. Gerok, and K. Decker.1990. Interleukin-6 release by rat liver macrophages. J. Hepatol.11:367-373.

11. Collart, M. A., P. Baeuerle, and P. Vassalli. 1990. Regulation oftumor necrosis factor alpha transcription in macrophages: in-volvement of four Kb-like motifs and of constitutive and induc-ible forms of NF-KB. Mol. Cell. Biol. 10:1498-1506.

12. Coutinho, A., and T. Meo. 1977. Genetic basis for unrespon-siveness to lipopolysaccharides in C57BL/10 Cr mice. Immuno-genetics 7:17-24.

13. Dignam, J. D., R. M. Lebowitz, and R. G. Roeder. 1983.Accurate transcription initiation by RNA polymerase II in asoluble extract from isolated mammalian nuclei. Nucleic AcidsRes. 11:1475-1489.

14. Dinarello, C. A. 1989. Interleukin 1 and its biologically relatedcytokines. Adv. Immunol. 44:153-205.

15. Dinarello, C. A., J. G. Cannon, S. M. Wolff, H. A. Bernheim, B.Beutler, A. Cerami, I. S. Figari, M. A. Palladino, and J. V.O'Connor. 1986. Tumor necrosis factor (cachectin) is an endog-enous pyrogen and induced production of interleukin 1. J. Exp.Med. 163:1433-1450.

16. Fischer, H., M. Dohlsten, U. Andersson, G. Hedlund, P. Erics-son, J. Hansson, and H. 0. Sjogren. 1990. Production of TNF-aand TNF-, by staphylococcal enterotoxin A activated human Tcells. J. Immunol. 144:4663-4669.

17. Fischer, H., M. Dohlsten, M. Lindvall, H.-O. Sjogren, and R.Carlsson. 1989. Binding of staphylococcal enterotoxin A toHLA-DR on B cell lines. J. Immunol. 142:3151-3157.

18. Flesch, I., B. Ecker, and E. Ferber. 1984. Acetyltransferasecatalyzed cleavage of arachidonic acid from phospholipids andtransfer to lysophosphatides in macrophages derived from bonemarrow. Eur. J. Biochem. 139:431-437.

19. Folks, T. M., K. A. Clouse, J. Justement, A. Rabson, E. Duh,J. H. Kehrl, and A. S. Fauci. 1989. Tumor necrosis factor et

induces expression of human immunodeficiency virus in achronically infected T-cell clone. Proc. Natl. Acad. Sci. USA86:2365-2368.

20. Fraser, J. D. 1989. High affinity binding of staphylococcalenterotoxin A and B to HLA-DR. Nature (London) 339:221-223.

21. Freudenberg, M. A., and C. Galanos. 1991. Tumor necrosisfactor alpha mediates lethal activity of killed gram-negative andgram-positive bacteria in D-galactosamine-treated mice. Infect.Immun. 59:2110-2115.

22. Freudenberg, M. A., D. Keppler, and C. Galanos. 1986. Require-ment for lipopolysaccharide-responsive macrophages in galac-tosamine-induced sensitization to endotoxin. Infect. Immun.51:891-895.

23. Galanos, C., and M. A. Freudenberg. 1990. TNF mediatesendotoxin shock: the protective effects of antibodies and corti-sone, p. 187-193 In B. Bonavida and G. Granger (ed.), TNF:structure, mechanism of action, role in disease and therapy. S.Karger, Basel.

24. Galanos, C., and 0. Luederitz. 1975. Electrodialysis of lipopoly-saccharides and their conversion to uniform salt forms. Eur. J.Biochem. 54:603-610.

25. Galanos, C., 0. Luederitz, and 0. Westphal. 1969. A newmethod for the extraction of R-lipopolysaccharides. Eur. J.Biochem. 9:245-249.

26. Ghosh, S., A. M. Gifford, L. R. Riviere, P. Tempst, G. P. Nolan,and D. Baltimore. 1990. Cloning of the p50 DNA binding subunitof NF-KB: homology to rel and dorsal. Cell 62:1019-1029.

27. Greene, W. C., E. Boehnlein, and D. W. Ballard. 1989. HIV-1,HTLV-1 and normal T-cell growth: transcriptional strategiesand surprises. Immunol. Today 8:272-278.

28. Griffin, G. E., K. Leung, T. M. Folks, S. Kunkel, and G. J.Nabel. 1989. Activation of HIV gene expression during mono-cyte differentiation by induction of NF-KB. Nature (London)339:70-73.

29. Haskill, S., A. A. Beg, S. M. Tompkins, J. S. Morris, A. D.Yurochko, A. Sampson-Johannes, K. Mondal, P. Ralph, andA. S. Baldwin. 1991. Characterization of an immediate-earlygene induced in adherent monocytes that encodes IKB-likeactivity. Cell 65:1281-1289.

30. Hohmann, H.-P., R. Remy, B. Poschl, and A. P. G. M. van Loon.1990. Tumor necrosis factors-a and -P bind to the same twotypes of tumor necrosis factor receptors and maximally activatethe transcription factor NF-KB at low receptor occupancy andwithin minutes after receptor binding. J. Biol. Chem. 265:15183-15188.

31. Hohmann, H.-P., R. Remy, C. Scheidereit, and A. P. G. M. vanLoon. 1991. Maintenance of NF-KB activity is dependent onprotein synthesis and the continuous presence of externalstimuli. Mol. Cell. Biol. 11:259-266.

32. Ikejima, T., S. Okusawa, J. M. van der Meer, and C. A.Dinarello. 1988. Induction by toxic-shock-syndrome toxin 1 of acirculating tumor necrosis-like substance in rabbits and ofimmunoreactive tumor necrosis factor and interleukin-1 fromhuman mononuclear cells. J. Infect. Dis. 158:1017-1025.

33. Janeway, C. A., Jr., J. Yagi, P. J. Conrad, M. E. Katz, B. Jones,S. Vroegop, and S. Buxser. 1989. T-cell responses to MLs and tobacterial proteins that mimic its behavior. Immunol. Rev.107:61.

34. Jupin, C., S. Anderson, C. Damais, J. E. Alouf, and M. Parant.1988. Toxic shock syndrome toxin 1 as an inducer of humantumor necrosis factors and -y interferon. J. Exp. Med. 167:752-761.

35. Karck, U., T. Peters, and K. Decker. 1988. The release of tumornecrosis factor from endotoxin-stimulated rat Kupffer cells isregulated by prostagladin E2 and dexamethasone. J. Hepatol.7:352-361.

36. Kawakami, K., C. Scheidereit, and R. G. Roeder. 1988. Identi-fication and purification of a human immunoglobulin-enhancer-binding protein (NF-KB) that activates transcription from ahuman immunodeficiency virus type 1 promoter in vitro. Proc.Natl. Acad. Sci. USA 85:4700-4704.

37. Kieran, M., V. Blank, F. Logeat, J. Vandekerckhove, F. Lottspe-ich, 0. Le Bail, M. B. Urban, P. Kourilsky, P. A. Baeuerle, andA. Israel. 1990. The DNA binding subunit of NF-KB is identicalto factor KBF1 and homologous to the rel oncogene product.Cell 62:1007-1018.

38. Kloos, W. E., and J. H. Jorgensen. 1985. Staphylococci, p. 143.In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and H. J.Shadomy (ed.), Manual of clinical microbiology, 4th ed. Amer-ican Society for Microbiology, Washington, D.C.

39. Lenardo, M., J. W. Pierce, and D. Baltimore. 1987. Protein-binding sites in Ig gene enhancers determine transcriptionalactivity and inducibility. Science 236:1573-1577.

40. Lenardo, M. J., and D. Baltimore. 1989. NF-KB: a pleiotropicmediator of inducible and tissue-specific gene control. Cell58:227-229.

INFECT. IMMUN.

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

NF-KB ACTIVATION BY S. AUREUS 2015

41. Lowenthal, J. W., D. W. Ballard, H. Bogerd, E. Bohnlein, andW. C. Greene. 1989. Tumor necrosis factor-a activation of theIL-2 receptor-a gene involves the induction of KB-specific DNAbinding proteins. J. Immunol. 142:3121-3128.

42. Marrack, P., M. Blackman, E. Kushnir, and J. Kappler. 1990.The cytotoxicity of staphylococcal enterotoxin B in mice ismediated by T cells. J. Exp. Med. 171:455-464.

43. Meichle, A., S. Schutze, G. Hensel, D. Brunsing, and M. Kronke.1990. Protein kinase C-independent activation of the nuclearfactor KB by tumor necrosis factor. J. Biol. Chem. 265:8339-8343.

44. Metcalf, D. 1977. Techniques for the clonal culture of hemopoi-etic cells in semisolid medium. Recent results in cancer re-search. Springer-Verlag, Berlin.

45. Meyer, R., E. N. Hatada, H.-P. Hohmann, M. Haiker, C.Bartsch, U. Roethlisberger, H.-W. Lahm, E. J. Schlaeger,A. P. G. M. van Loon, and C. Scheidereit. 1991. Cloning of theDNA-binding subunit of human nuclear factor KB: the level ofits mRNA is strongly regulated by phorbol ester or tumornecrosis factor a. Proc. Natl. Acad. Sci. USA 88:966-970.

46. Mollick, J. A., R. C. Cook, and R. R. Rich. 1989. Class II MHCmolecules are specific receptors for Staphylococcus enterotoxinA. Science 244:817-820.

47. Morrison, D. C., and J. L. Ryan. 1987. Endotoxins and diseasemechanisms. Annu. Rev. Med. 38:417-432.

48. Mourad, W., R. S. Geha, and T. Chatila. 1990. Engagement ofmajor histocompatibility complex class II molecules inducessustained, lymphocyte function-associated molecules 1-depen-dent cell adhesion. J. Exp. Med. 172:1513-1516.

49. Movat, H. Z., M. I. Cybulsky, I. G. Colditz, M. K. W. Chan, andC. A. Dinarello. 1987. Acute inflammation in gram-negativeinfection: endotoxin, interleukin 1, tumor necrosis factor andneutrophils. Fed. Proc. 46:97-104.

50. Nabel, G., and D. Baltimore. 1987. An inducible transcriptionfactor activates expression of human immunodeficiency virus inT cells. Nature (London) 326:711-713.

51. Nolan, G. P., S. Gosh, H.-L. Liou, T. Tempst, and D. Baltimore.1991. DNA binding and IKB inhibition of the cloned p65 subunitof NF-KB, a rel-related polypeptide. Cell 64:961-969.

52. Osborn, L., S. Kunkel, and G. J. Nabel. 1989. Tumor necrosisfactor a and interleukin 1 stimulate the human immunodefi-ciency virus enhancer by activation of the nuclear factor KB.Proc. Natl. Acad. Sci. USA 86:2336-2340.

53. Parsonnet, J., and Z. A. Gillis. 1988. Production of tumornecrosis factor by human monocytes in response to toxic-shock-syndrome toxin-1. J. Infect. Dis. 158:1026-1033.

54. Pointdexter, N. J., and P. M. Schlievert. 1985. Toxic-shock-toxin-i-induced proliferation of lymphocytes: comparison of themitogenic response of human, murine, and rabbit lymphocytes.J. Infect. Dis. 151:65.

55. Ruben, S. M., P. J. Dillon, R. Schreck, T. Henkel, C.-H. Chen,M. Maher, P. M. Baeuerle, and C. A. Rosen. 1991. Isolation ofa rel-related human cDNA that potentially encodes the p65-kDsubunit of NF-KB. Science 251:1490-1493.

56. Scholl, P., A. Diez, J. Mourad, J. Parsonnet, R. S. Geha, and T.Chatila. 1989. Toxic shock syndrome toxin 1 binds to majorhistocompatibility complex class II molecules. Proc. Natl.Acad. Sci. USA 86:4210.

57. Sen, R., and D. Baltimore. 1986. Inducibility of K immunoglob-ulin enhancer-binding protein NF-KB by a posttranslationalmechanism. Cell 47:921-928.

58. Sen, R., and D. Baltimore. 1986. Multiple nuclear factorsinteract with the immunoglobulin enhancer sequences. Cell46:705-716.

59. Tracey, J. J., B. Beutler, F. Lowry, J. Merryweather, S. Wolpe,M. Milsark, R. J. Hariri, T. J. Fahey, J. D. Zentella, G. T.Shires, and A. Cerami. 1986. Shock and tissue injury induced byrecombinant human cachectin. Science 234:470-474.

60. Uchiama, G., Y. Kamagata, X.-J. Yan, M. Kohno, M. Yoshi-oka, H. Fujikawa, H. Igarashi, M. Okubo, F. Awano, T. Saito-Taki, and M. Nakono. 1987. Study of the biological activities oftoxic shock syndrome toxin-1. II. Induction of the proliferativeresponse and the interleukin-2 production by T-cells fromhuman peripheral blood mononuclear cells stimulated with thetoxin. Clin. Exp. Immunol. 68:638.

61. Visvanathan, K. V., and S. Goodbourn. 1989. Double-strandedRNA activates binding of NF-KB to an inducible element in thehuman p-interferon promoter. EMBO J. 8:1129-1138.

62. Westphal, O., 0. Luederitz, C. Galanos, H. Mayer, and E. T.Rietschel. 1985. The story of bacterial endotoxin, p. 13-34. InL. Chedid, J. W. Hadden, F. Spreafico, P. Dukor, and D.Willoughby (ed.), Advances in immunopharmacology. Proceed-ings of the Third International Conference on Immunopharma-cology. Florence, Italy, 6 to 9 May 1985. Pergamon Press, NewYork.

VOL. 60, 1992

on February 13, 2020 by guest

http://iai.asm.org/

Dow

nloaded from